I Didn't pay off my mortgage, and I feel FUCKING great!

[FIreDrill]

Okay, decided to run the numbers over the super long term. Imputed what the interest rate for a 30 year mortgage would be from the interest rate on a 10 year treasury using the formula provided here. All values assume a starting mortgage of $100k and $2k of monthly income that can either be thrown entirely at the mortgage until it is gone, and then invested in stocks, or used to make the minimum monthly payment on the mortgage and invest the difference in the stock market.



Out of 1391 start months with at least 30 years of stock market data, paying off your mortgage was the better outcome in 175 months (12.6% of the time). The average benefit of making only minimum mortgage payments for 30 years was ~$122k (in dollars inflation adjusted to the start of the 30 year period), and the median benefit was ~$82,700.

Adding in the assumption that our simulated person had already exceeded their standard deduction, that mortgage interest was tax deductible, and that their marginal tax rate was 25% produced more favorable results.

Wow, awesome info!  Thanks for taking the time to put this together.

I wonder what the numbers would turn out like for mortgages of 200k, 300k, and so forth.  I'm guessing the final results would have an exponential effect rather than linear due to compounding over a 30 year period.

It would be really cool to have a tool available that you could plug in specific numbers for.  Like interest rate, mortgage term, pre-tax investment savings and mortgage interest tax savings over different brackets.  I bet a lot of people In the MMM community would benefit from such a tool.

Sent from my SM-G935F using Tapatalk

[Dicey]

Oh, maizeman, I loves me some mortgage porn! You're the bomb!

[maizefolk]

Diane C, glad to hear the analysis was a hit! FrozenBits, I would guess that as long as the ratio of mortgage to assumed monthly income stays the same, the change would just be linear, not exponential, but it would be easy enough to test.

Right now the script doesn't factor in any pre-tax savings space forgone because the money that would have been put into the 401k/IRA/403b etc went to pay off the mortgage. Again, it'd be a relatively straightforward addition, but then I start having to make even more assumptions about marginal tax rates both before and after retirement.

[brooklynguy]

Maizeman - thanks for the excellent graphical analysis (as usual).

Imputed what the interest rate for a 30 year mortgage would be from the interest rate on a 10 year treasury using the formula provided here.

In the forum's chief mortgage thread, starting at reply # 310, we had an extended philosophical debate over whether, for purposes of a current mortgage loan borrower running the "to prepay or not to prepay" cost-benefit analysis on the basis of historical stock market returns, it is more appropriate to use today's (historically low) mortgage interest rates or actual then-prevailing historical mortgage interest rates.

It would be really cool to have a tool available that you could plug in specific numbers for. 

You can use cFIREsim for the purpose of determining the performance of a leveraged-investing-via-mortgage strategy using actual historical market returns and the inputs and assumptions of your choice, as we did in the thread cited above.  In particular, see the original post and reply # 291 in that thread.

[RoostKing]

My critique/counter-argument is limited to those who are doing it only for the psychological reason of owning a property "free and clear".   And even for those folks- you are still looking at a guaranteed return of whatever your interest rate is (minus any ability to itemize interest), so don't want to come off as being anything but encouraging, as this is still a good way to build wealth over time with reduced risk.

Yep.

The big issue I have against many of those that are doing it for emotional reasons is that seems to go against the MMM philosophy of optimization and the ability to go against "normal" emotions in order make your stache grow as quickly as possible.  Like not being a consumer sucker.  Granted this is a much more complex topic than standard consumerism but it can also result in the loss of hundreds of thousands of dollars over a 30 year period. 

I believe MMM himself has a paid off home, but I wonder how his personal tax liability would change if he mortgaged the home and put a couple hundred grand into his betterment account where he could take advantage of tax loss harvesting in addition to the tax savings from mortgage interest deductions as well.

In the big picture, we are looking at 2 great options.  One is better and the other is betterest :p.  Although, my minds fascination with optimization drives me to find which one is which.


Sent from my SM-G935F using Tapatalk

This is the category I fall in to, and I am sorta regretting it. I had a high paying job overseas that had a definitive end date, so I put as much towards my house as possible (2 years ago, I put $65,000 down on a $236,000 loan, 3.65%) I didnt quite pay it off, but at normal payments, I will have it paid off in a little over 4 years, which is great, but I think of what I have missed out on had the money been in the market the past couple of years. On the flip side, what if the market would have tanked, I would not have been able to recoup that high $$$$ i was making. Hind sight is 20/20.

[maizefolk]

Thanks Brooklynguy! And as usual, you have linked to an extremely intriguing and in depth discussion of this topic that I completely missed the first time around.

Of the two options it seems pretty clear using the interest rates in the time periods in question (which will almost always be higher than today) would be more conservative while that using today's (well 2015's) low interest rates would be more likely to favor maintaining leverage by paying off the mortgage.

When given two options like this where it's not clear which is a better fit for reality, my gut instinct is usually to pick whichever one is going to bias things more AGAINST my expected outcome, because it provides one less confounding variable to argue about if I do end up get a result consistent with my expected outcome.

Although, just eyeballing the graphs -- I'm not on my home computer so cannot run the numbers right now -- it looks like variation in stock market performance has a much bigger impact on whether it makes sense to pay off your mortgage quickly than the interest rate on the mortgage itself.

[FIreDrill]

This is the category I fall in to, and I am sorta regretting it. I had a high paying job overseas that had a definitive end date, so I put as much towards my house as possible (2 years ago, I put $65,000 down on a $236,000 loan, 3.65%) I didnt quite pay it off, but at normal payments, I will have it paid off in a little over 4 years, which is great, but I think of what I have missed out on had the money been in the market the past couple of years. On the flip side, what if the market would have tanked, I would not have been able to recoup that high $$$$ i was making. Hind sight is 20/20.

Yes, if only we all knew what was going to happen in the future ;).  This really leaves us with making our best educated guess according to historical information.

We may be coming into a situation similar to yours with unstable income.  Our plan is to accumulate a large cash/bond buffer (1 year bare bones expenses) without pre-paying our mortgage and then continue to invest like mad.

Hopefully this gives us enough flexibility to weather any storms that come our way while still getting the most out of investing in the stock market.  But until we know for certain what is happening, we will continue to go 100% stocks.

Sent from my SM-G935F using Tapatalk

[brooklynguy]

Of the two options it seems pretty clear using the interest rates in the time periods in question (which will almost always be higher than today) would be more conservative while that using today's (well 2015's) low interest rates would be more likely to favor maintaining leverage by paying off the mortgage.

Yes, that's definitely true.  But, as I argued in the other thread, that option is also inconsistent with the Bengen/Trinity-study-style history-based SWR retirement planning we typically do around here, which tests how portfolios subjected to assumed future spending plans (in our case, the known future spending plan that matches the mortgage loan's (non-inflation adjusted) amortization schedule) would have fared historically using actual historical market performance.

[Lmoot]

I still feel like comparing the market to primary residence is not an equal comparison. Most people buy into the market with the only goal of making money. With a primary residence, people are more likely to spend more capital on items that don't improve (and can actually deduct from) returns. Comparing the market to investment properties is more accurate IMO. Though I do realize the very topic of discussion is whether or not to pay off a primary mortgage. But still...putting something that is clearly designed as an investment as its only goal, up against something that can turn out to be an investment by default (but not necessarily design), is a little "unfair". But it also makes it that much more impressive, if going by the chart above, that housing can be a more profitable investment without even trying. Can you imagine if you tried? That's  one reason why I got into investment properties. More cashflow for today, and asset value for tomorrow. Best of both worlds.

[maizefolk]

Hi Lmoot. I guess the question is, if you're going to buy a house anyway, what's the most economically efficient way to do so. So I see this less as a question about investments, and more about the general lifestyle optimization which is also a big part of the MMM community.

Brooklynguy, yup, I'd agree it's a different and inconsistent assumption from what we use to model SWRs from portfolios. But I don't think the assumption that any point in recorded stock history is equally likely to be representative of returns from today going forward is used in Trinity studies because it's particularly likely to be correct, it's just that we have so little data (less than 4 non-overlapping 30 year intervals) that as soon as you start trying to control for other variables you run a substantial risk of overfitting your model and ending up worse off than when you started.

I guess the philosophical question is, if we had enough data (say a couple of thousand years instead of a bit more than a century) of stock market and interest rate data, would you want to try to take interest rates into account in determining that a safe or sustainable withdrawal rate from a stock portfolio would be?

[FIreDrill]

Hi Lmoot. I guess the question is, if you're going to buy a house anyway, what's the most economically efficient way to do so. So I see this less as a question about investments, and more about the general lifestyle optimization which is also a big part of the MMM community.

Agreed, in this scenario we would be comparing guaranteed mortgage payoff returns to stock market or real estate investment returns.  If you mortgaged a rental property for 70% LTV  and then purchase 2 additional rental properties with a 70% LTV mortgage that have good cash flow yielding 10-15% after expenses, you would just be changing the investment your using the debt for.  Which is real estate instead of the stock market.

Although you would be taking on significant leverage at this point compared to investing in the stock market.  Investing in real estate can get a little ridiculous but be very profitable if done correctly.

Sent from my SM-G935F using Tapatalk

[brooklynguy]

But I don't think the assumption that any point in recorded stock history is equally likely to be representative of returns from today going forward is used in Trinity studies because it's particularly likely to be correct, it's just that we have so little data (less than 4 non-overlapping 30 year intervals) that as soon as you start trying to control for other variables you run a substantial risk of overfitting your model and ending up worse off than when you started.

Indeed, Trinity-style retirement planning does implicitly recognize the incorrectness of this assumption in reality.  When we use historical success rates as a proxy for likelihood of future success (which assumes, among other assumptions, that "any point in recorded stock history is equally likely to be representative of returns from today going forward"), we implicitly recognize the falsity of that assumption by selecting as our target an exceedingly high likelihood of success (usually 95%+).  So the "worst case" or "near worst case" nature of this type of history-based planning compensates for its willful blindness to the unequal market conditions in existence at all the various period start dates.  If we had, and utilized, a better future-returns-predictor (whether via a larger historical data set, the incorporation of valuation-based metrics like CAPE, or something else), then it would no longer make as much sense to plan for the future on the basis of history's worst case scenarios.

[FIreDrill]

A couple months later and we are crossing the 250k invested mark this week, which surpasses the full value of our house.  So far we are on track to break 300k invested by years end.

Boom!

Sent from my SM-G935F using Tapatalk

[Dicey]

A couple months later and we are crossing the 250k invested mark this week, which surpasses the full value of our house.  So far we are on track to break 300k invested by years end.

Boom!

Sent from my SM-G935F using Tapatalk

Boom yeah!

[elysianfields]

Okay, decided to run the numbers over the super long term. Imputed what the interest rate for a 30 year mortgage would be from the interest rate on a 10 year treasury using the formula provided here. All values assume a starting mortgage of $100k and $2k of monthly income that can either be thrown entirely at the mortgage until it is gone, and then invested in stocks, or used to make the minimum monthly payment on the mortgage and invest the difference in the stock market.



Out of 1391 start months with at least 30 years of stock market data, paying off your mortgage was the better outcome in 175 months (12.6% of the time). The average benefit of making only minimum mortgage payments for 30 years was ~$122k (in dollars inflation adjusted to the start of the 30 year period), and the median benefit was ~$82,700.

Adding in the assumption that our simulated person had already exceeded their standard deduction, that mortgage interest was tax deductible, and that their marginal tax rate was 25% produced more favorable results.

Nice graph.  Alas, you've made two, or maybe two and one-half enormous errors in your assumptions:

1.  The federal income tax has only existed since 1913.  There was nothing to deduct mortgage interest against before 1913.  Furthermore, at the beginning of the income tax, it was only paid by the very rich; only during World War II was the federal income tax democratized to the middle class through the implementation of federal income tax withholding.
2.  The mortgage market for residential homes changed drastically with the National Housing Act of 1934, which created the FHA and FSLIC and encouraged longer-duration  (15 to 30 year) amortizing mortgages.  Previously, most mortgages were for smaller LTV amounts (usually capped at 50%), and were five-year interest-only balloon mortgages.
2 1/2.  The mortgage market changed dramatically again with the creation of the Federal National Mortgage Association and the securitization of mortgages.  This brought more lenders into the market and improved the liquidity of mortgages; it also created a standardized mortgage in the form of the "conforming loan".

For these reasons, I wouldn't read too much into your data or graph until at least 1945, if not later.

[brooklynguy]

Alas, you've made two, or maybe two and one-half enormous errors in your assumptions

These are no more "errors" than is the common practice in Trinity-style methodology of assuming that historical investors invested using investment vehicles that, in reality, would have been unavailable to them (like index funds).  The purpose of these types of back-tests is not to determine how actual investors in the past would have actually fared subject to actual then-applicable real-world constraints; instead, it's to determine how hypothetical investors in the past would have fared (using actual historical market performance) if it were somehow possible for them to invest using the investment options available today.  Of course, this does raise philosophical questions about the appropriateness (and usefulness) of the entire back-testing exercise -- see the discussion on that point in the handful of posts immediately above, and in the mortgage thread cited in reply # 53.

[elysianfields]

These are no more "errors" than is the common practice in Trinity-style methodology of assuming that historical investors invested using investment vehicles that, in reality, would have been unavailable to them (like index funds).  The purpose of these types of back-tests is not to determine how actual investors in the past would have actually fared subject to actual then-applicable real-world constraints; instead, it's to determine how hypothetical investors in the past would have fared (using actual historical market performance) if it were somehow possible for them to invest using the investment options available today.  Of course, this does raise philosophical questions about the appropriateness (and usefulness) of the entire back-testing exercise -- see the discussion on that point in the handful of posts immediately above, and in the mortgage thread cited in reply # 53.

If I understood correctly, the graph attempts to depict not just market performance, but a comparison of "investing" marginal dollars into paying down a mortgage instead of investing in the stock market.

While historical stock market return data has its own issues, as you mention, I'm not picking bones with historical stock market performance in isolation, but rather in this context where that data is graphed against the ten-year Treasury yield as a proxy for historical mortgage data.

The graph depicts, in essence, stock returns versus bond returns, which isn't what it purports to depict.

In fact, I would argue that the multiple options currently available in the mortgage market, combined with the changes in mortgage loans seen in the 20th century, makes the attempt to reduce the comparison of investing the marginal dollar (in the market versus a mortgage) impossible to graph reasonably.

While US mortgage interest rates move in tandem with the 10-year Treasury, 15-year and 30-year fixed-rate mortgages nearly always sport significantly higher interest rates because of the length of the rate lock.  Because of this, the 30-year Treasury note would make a better proxy for the 30-year fixed-rate mortgage.  Moreover, as I mentioned, 15- and 30-year conforming loans didn't exist before the 1950s.  Therefore using the 10-year as a proxy doesn't work well at all.  I suppose you could argue that it's better than nothing, but then I would say you shouldn't put much faith in such a poor model.

Note, by the way, that I'm arguing against my own actions (I carry a mortgage when I could easily pay it off), because the 10-year Treasury is easier to beat that the higher 30-year fixed-rate mortgage I'm paying.

[brooklynguy]

First, the graph doesn't use the 10-year Treasury yield as a proxy for historical mortgage interest rates -- it uses the results of the formula cited in maizeman's original post which includes the 10-year Treasury yield as one of its component variables.

But that's beside the point I was trying to make, which is that the real world impossibility of the existence of assumed historical conditions is not a flaw (or at least, if it is a flaw, is a known and accepted flaw) in these types of backtests.  We can argue about which statistic best serves a proxy for historical interest rates on 30-year quasi-government subsidized fixed-rate mortgage loans during periods in which such loans did not actually exist, but the fact that they did not actually exist does not create an "error" in the backtest in the same way that the nonexistence of cheap index funds for most of the stock market's history does not create an error in a backtest that purports to gauge the historical performance of an investment strategy employing such investment vehicles.

[elysianfields]

First, the graph doesn't use the 10-year Treasury yield as a proxy for historical mortgage interest rates -- it uses the results of the formula cited in maizeman's original post which includes the 10-year Treasury yield as one of its component variables.

Point well taken.  I missed that.  My apologies for the oversight.

But that's beside the point I was trying to make, which is that the real world impossibility of the existence of assumed historical conditions is not a flaw (or at least, if it is a flaw, is a known and accepted flaw) in these types of backtests.  We can argue about which statistic best serves a proxy for historical interest rates on 30-year quasi-government subsidized fixed-rate mortgage loans during periods in which such loans did not actually exist, but the fact that they did not actually exist does not create an "error" in the backtest in the same way that the nonexistence of cheap index funds for most of the stock market's history does not create an error in a backtest that purports to gauge the historical performance of an investment strategy employing such investment vehicles.

Perhaps "error" isn't the correct term for flawed data.  In any case, even with your assumptions, the applicability of a dataset, a large proportion of which is drawn from such a different period of financial and fiscal history, to a person deciding today, under current conditions, where to invest their marginal dollar of savings, is at best questionable.

[yachi]

Out of 1391 start months with at least 30 years of stock market data, paying off your mortgage was the better outcome in 175 months (12.6% of the time).

That's an amazingly bad track record of being the better outcome.  I can't tell, is your comparison actually varying the mortgage interest rate?  Baring relocating, my mortage rate would not rise over the next 30 years even if mortgage rates for new mortgages increase.  It seems today's interest rates are so low as to be certain to fall in months where paying off the mortgage would be the worse outcome.

[maizefolk]

Out of 1391 start months with at least 30 years of stock market data, paying off your mortgage was the better outcome in 175 months (12.6% of the time).

That's an amazingly bad track record of being the better outcome.  I can't tell, is your comparison actually varying the mortgage interest rate?  Baring relocating, my mortage rate would not rise over the next 30 years even if mortgage rates for new mortgages increase.  It seems today's interest rates are so low as to be certain to fall in months where paying off the mortgage would be the worse outcome.

Yes, the comparison is trying to estimate what the interest rate would have been on a conventional 30 year mortgage in each start month if such a mortgage had been available at that point in history. So the interest rate on the hypothetical mortgage for the 30 years of the simulation is set based on our estimate of what a 30 year mortgage interest rate would have been in the first month of that 30 year interval.*

There are essentially two ways for paying off the mortgage to be the better option: 1) extremely high interest rate on the mortgage, but declining interest rates/inflation after you take the mortgage out 2) severe and prolonged stock market crash (which certanly can happen).

*EF, this would be my answer to your concern that I'm essentially just comparing bond and stock yields. Estimating returns on bonds tend to work with "constant maturity bonds" and both the value of the investment and the rate of require move with changes in interest rates throughout the simulated investment interval. A mortgage really is different in that you're locking in one interest rate for an entire 30 year interval, so it behaves very differently from a bond, even though I'm using interest rate data in the starting month to estimate what interest rate a mortgage would have (and the model used does appear to have a pretty good R^2 value with actual 30 year mortgage rates in the interval where modern 30 year mortgages have existed).

[dude]

Another member of the not-paying-the-mortgage off crowd.  At 3.25% fixed, and a payment that is far below prevailing rental rates in our area, I consider my mortgage payment an extremely good deal.  As we've serially refi'ed down to this rate (from an initial 80/15 loan at 6%/8%), we've plowed the mortgage savings into our investment accounts.  We currently owe $392k on the mortgage, home equity stands at around $300k, and investment accounts are at $954k. In addition, we own a second home outright (wife's parent's home), that we will liquidate upon their eventual passing; the proceeds could then be used to pay down our mortgage, or better yet, we can invest the proceeds and pay our mortgage with the returns. At 3.25%, our interest payments are down considerably and equity is building pretty fast.  Another consideration in all of this is I'll have an inflation-adjusted pension coming my way in 2 years when I retire.  So having a mortgage in FIRE works fine for me.

[albireo13]

Don't forget those savings are pre-tax so factor the tax hit when comparing against outstanding mortgage.
This is because you pay your mortgage bill with after-tax dollars.

[runewell]

Don't forget those savings are pre-tax so factor the tax hit when comparing against outstanding mortgage.
This is because you pay your mortgage bill with after-tax dollars.

When I do my analysis I am not assuming that my savings is going into some tax-deferred retirement plan, I treat the money as it's a taxable investment in an attempt to be as apples-to-apples as possible. 

[Dicey]

Another member of the not-paying-the-mortgage off crowd.  At 3.25% fixed, and a payment that is far below prevailing rental rates in our area, I consider my mortgage payment an extremely good deal... We currently owe $392k on the mortgage, home equity stands at around $300k, and investment accounts are at $954k

Dude, welcome to the club! Nice to have another voice in the choir!

[FIreDrill]

Another member of the not-paying-the-mortgage off crowd.  At 3.25% fixed, and a payment that is far below prevailing rental rates in our area, I consider my mortgage payment an extremely good deal... We currently owe $392k on the mortgage, home equity stands at around $300k, and investment accounts are at $954k

Dude, welcome to the club! Nice to have another voice in the choir!

Yes! Welcome!

Seems like the choir has been growing at a decent rate here recently ;)



Sent from my SM-G935F using Tapatalk